Conversational Interface Rule Segmentation for Response Adaptability

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conversational computing interfaces that rely on hard-coded skills are limited in performing complex or novel behaviors and cannot produce response utterances that vary based on different outcomes from executing a computer-executable plan.

Innovation Solution

A method where a conversational computing interface selects an applicable generation rule, passes parameters to additional rules, and recursively applies these rules to extend the computer-executable plan, generating candidate responses that can vary in format and content based on outcomes, using a trained generation model to produce descriptive content related to the plan's results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If hard-coded skills are used in conversational computing interfaces, then the system can perform predefined tasks, but it cannot handle complex or novel behaviors and produces limited response variations

Engineering Contradiction:

Improveability to handle novel situationsVSAvoidsystem architecture complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system segments response generation into multiple independent generation rules, each handling specific aspects of response creation. This allows the system to handle complex behaviors by combining simple, modular rules rather than using a monolithic hard-coded skill system.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system dynamically selects and applies generation rules based on the conversational event and executed plan, rather than using static hard-coded responses. This dynamic rule application enables the system to adapt to novel situations while maintaining manageable complexity through algorithmic control.

Inventive Principle:

Principle #15Dynamics

2Adaptability or versatility

If hard-coded skills are used, then the system structure is simple, but the response utterances cannot vary based on different outcomes from executing a computer-executable plan

Engineering Contradiction:

Improveresponse variation capabilityVSAvoidsystem development ease

Core Design Contradiction:

Adaptability or versatilityVSEase of manufacture

Solution Approach 1:

The generation rules are designed to be universal and reusable across different conversational events and plan outcomes. A single rule can handle multiple scenarios by parameterizing its behavior, reducing the need to create separate hard-coded responses for each situation while maintaining response variation capability.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system changes parameters of generation rules dynamically based on plan outcomes and conversational context. By modifying rule parameters rather than creating entirely new hard-coded responses, the system achieves response variation while keeping development easier through parameterization rather than comprehensive rule creation.

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If a trained generation model is used to produce descriptive content, then the system can generate flexible and varied response utterances, but the computational resources and processing time increase

Engineering Contradiction:

Improveresponse generation flexibilityVSAvoidcomputational energy consumption

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary actions by pre-defining generation rules and their parameters during system setup or training. This preliminary configuration allows the actual response generation to be more efficient, as the model only needs to apply pre-established rules rather than learn and generate from scratch for each situation, reducing real-time computational energy consumption.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11790897B2Response generation for conversational computing interface

Publication Date: 2023.10.17 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

A computer-implemented method of responding to a conversational event is presented. The method comprises receiving a conversational event at a conversational computing interface. Based on the received conversational event, an applicable generation rule of a plurality of candidate generation rules is selected. The applicable generation rule is configured with one or more parameters. A computer-executable plan is then selected based on the selected generation rule. The one or more parameters are passed from the selected generation rule to one or more additional generation rules. The one or more additional generation rules configured with the one or more parameters are recursively applied to extend the selected computer-executable plan. One or more candidate responses to the conversational event are output via the conversational computing interface based on the recursive application of the one or more additional generation rules configured with the one or more parameters.